To whomever can help:
I'm studying spanning-tree Protocol and I'm not clear as to why spanning tree was implemented as it was.
Why cant each switch have a topological map and decide the best path for itself like a router does? is this not possible? is it because layer 2 is only concerned with getting the frame between each individual link? I do not understand why all traffic must go to the root bridge before being forwarded. My mentality is, "think for yourself," so I don't understand why there must be a hierarchical design. to me it does not make sense to send all traffic to one swtich for it to then turn around and forward it to the correct place instead of the first switch the frame hits just sending it straight to the destination provided that the mac table is updated.
I think my main problem is still why do we have to have layer 2 and layer 3? why not just integrate those 2 parts? I think the OSI model should only include application presentation session transport pathing (layers 2 and 3 put together) and physical layer. I thnink that IPv6 does partially address this problem but it won't be globally/universally adopted by busiensses of all sizes and home users for years to come. not until the ISPs completely change over.
The bottom line is I don't understand Spanning tree and why the routers can't, "see" the topology of the layer 2 network, think for themselves, and act as an intellegent individual in the group. for example, I say, "hey bob, could you pass this memo to alice and don't give it to anyone else or let her give it to anyone else." But instead it works as if Bob doesn't understand the request and photocopies the memo and passes it to everyone and they all don't want it because it's not for them. If one uses people as an analogy for networking equipment, it's easy to see how it breaks down. If bob copies and distributes a confidential memo, not only does it compromise the secret nature of the memo but it also wastes the time of all of the people who got it and therefore decreases productivity. In this scenario Bob would be repremanded or fired. So why do we let switches do such things with our data? If the swtiches could understand me, I'd say to them, "Stop being so inefficient, wasting everyone's time, and think for yourself. Look at the network find the best path for yourself, send the frame that way and make sure it gets there without needlessly bothering uninvolved parties, who don't need to know anyway."
I hope this helps you to help me. Thanks in advance.
Please do not double post.
Try reading wiki see http://en.wikipedia.org/wiki/Spanning_Tree_Protocol
When they wrote STP in 1985 things were slightly different. We did not have switches only bridges which was normally a PC with two interfaces.
IP was still hiding. Appletalk Token Ring IPX/SPX were the technology of the day. It is now 26 years later and you will learn the are a few go fast versions commonly used.
what happens when switch receives L3 broadcast, it will forward it on all ports;
if STP is disabled, switches will keep forwarding the same b-cast back and forth;
with STP on, there is no redundant paths although you could see one with all cables connected in such way;
STP eliminates loops;
switches have MAC to Port and VLANs tables; routers have routing tables and ARP table; PC builds ARP table;
I'm Sorry. I didn't know where to post and figured like other places that you have to post, post, and post some more to get heard by the community. Thank you for a quick reply. This was my first post so I thought that it would take weeks if not months before I got any reply at all. I have these ideas and I think it would be great to try to implement them because of all this redundant traffic that seems to hamper performance. "why can't it work better, faster, cheaper, and more efficiently?" that's my motto. I think STP is due for an overhaul. There's no reason that we should still be using 1980's technology. In my humble opinion of course! thanks for the great resources.
i will recommend please select a book and go through it.